American Journal of Epidemiology - Advance Access

Health Behaviors From Early to Late Midlife as Predictors of Cognitive Function: The Whitehall II Study

2009-07-02

The authors examined associations of health behaviors over a 17-year period, separately and in combination, with cognition in late midlife in 5,123 men and women from the Whitehall II study (United Kingdom). Health behaviors were assessed in early midlife (mean age = 44 years; phase 1, 1985–1988), in midlife (mean age = 56 years; phase 5, 1997–1999), and in late midlife (mean age = 61 years; phase 7, 2002–2004). A score of the number of unhealthy behaviors (smoking, alcohol abstinence, low physical activity, and low fruit and vegetable consumption) was defined as ranging from 0 to 4. Poor (defined as scores in the worst sex-specific quintile) executive function and memory in late midlife (phase 7) were analyzed as outcomes. Compared with those with no unhealthy behaviors, those with 3–4 unhealthy behaviors at phase 1 (odds ratio (OR) = 1.84, 95% confidence interval (CI): 1.27, 2.65), phase 5 (OR = 2.38, 95% CI: 1.76, 3.22), and phase 7 (OR = 2.76, 95% CI: 2.04, 3.73) were more likely to have poor executive function. A similar association was observed for memory. The odds of poor executive function and memory were the greater the more times the participant reported unhealthy behaviors over the 3 phases. This study suggests that both the number of unhealthy behaviors and their duration are associated with subsequent cognitive function in later life.

TWO AUTHORS REPLY

Harder, T., Plagemann, A. — 2009-07-01

RE: "BIRTH WEIGHT, EARLY WEIGHT GAIN, AND SUBSEQUENT RISK OF TYPE 1 DIABETES: SYSTEMATIC REVIEW AND META-ANALYSIS"

Cardwell, C. R., Patterson, C. C. — 2009-07-01

Identifying a National Death Index Match

Fillenbaum, G. G., Burchett, B. M., Blazer, D. G. — 2009-06-30

Data from the National Death Index (NDI) are frequently used to determine survival status in epidemiologic or clinical studies. On the basis of selected information submitted by the investigator, NDI returns a file containing a set of candidate matches. Although NDI deems some matches as perfect, multiple candidate matches may be available for other cases. Working across data from the Duke University site of the Established Populations for Epidemiologic Studies of the Elderly (EPESE), NDI, and the Social Security Death Index (SSDI), the authors found that, for this Established Populations for Epidemiologic Studies of the Elderly cohort of 1,896 cases born before 1922 and alive as of January 1, 1999, a match on Social Security number plus additional personal information (specific combinations of last name, first name, month of birth, day of birth) resulted in agreement between NDI and Social Security Death Index dates of death 94.7% of the time, while comparable agreement was found for only 12.3% of candidate decedents who did not have the required combination of information. Thus, an easy to apply algorithm facilitates accurate identification of NDI matches.

The Physical and Mental Health of Australian Vietnam Veterans 3 Decades After the War and Its Relation to Military Service, Combat, and Post-Traumatic Stress Disorder

O'Toole, B. I., Catts, S. V., Outram, S., Pierse, K. R., Cockburn, J. — 2009-06-29

The long-term health consequences of war service remain unclear, despite burgeoning scientific interest. A longitudinal cohort study of a random sample of Australian Vietnam veterans was designed to assess veterans’ postwar physical and mental health 36 years after the war (2005–2006) and to examine its relation to Army service, combat, and post-traumatic stress disorder (PTSD) assessed 14 years previously (1990–1993). Prevalences in veterans (n = 450) were compared with those in the Australian general population. Veterans’ Army service and data from the first assessments were evaluated using multivariate logistic regression prediction modeling. Veterans’ general health and some health risk factors were poorer and medical consultation rates were higher than Australian population expectations. Of 67 long-term conditions, the prevalences of 47 were higher and the prevalences of 4 were lower when compared with population expectations. Half of all veterans took some form of medication for mental well-being. The prevalence of psychiatric diagnoses exceeded Australian population expectations. Military and war service characteristics and age were the most frequent predictors of physical health endpoints, while PTSD was most strongly associated with psychiatric diagnoses. Draftees had better physical health than regular enlistees but no better mental health. Army service and war-related PTSD are associated with risk of illness in later life among Australian Vietnam veterans.

Are Americans Feeling Less Healthy? The Puzzle of Trends in Self-rated Health

Salomon, J. A., Nordhagen, S., Oza, S., Murray, C. J. L. — 2009-06-29

Although self-rated health is proposed for use in public health monitoring, previous reports on US levels and trends in self-rated health have shown ambiguous results. This study presents a comprehensive comparative analysis of responses to a common self-rated health question in 4 national surveys from 1971 to 2007: the National Health and Nutrition Examination Survey, Behavioral Risk Factor Surveillance System, National Health Interview Survey, and Current Population Survey. In addition to variation in the levels of self-rated health across surveys, striking discrepancies in time trends were observed. Whereas data from the Behavioral Risk Factor Surveillance System demonstrate that Americans were increasingly likely to report "fair" or "poor" health over the last decade, those from the Current Population Survey indicate the opposite trend. Subgroup analyses revealed that the greatest inconsistencies were among young respondents, Hispanics, and those without a high school education. Trends in "fair" or "poor" ratings were more inconsistent than trends in "excellent" ratings. The observed discrepancies elude simple explanations but suggest that self-rated health may be unsuitable for monitoring changes in population health over time. Analyses of socioeconomic disparities that use self-rated health may be particularly vulnerable to comparability problems, as inconsistencies are most pronounced among the lowest education group. More work is urgently needed on robust and comparable approaches to tracking population health.

Mortality Rates Among Trichlorophenol Workers With Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin

Collins, J. J., Bodner, K., Aylward, L. L., Wilken, M., Bodnar, C. M. — 2009-06-26

The authors examined 1,615 workers exposed to dioxins in trichlorophenol production in Midland, Michigan, to determine if there were increased mortality rates from exposure. Historical dioxin levels were estimated by a serum survey of workers. Vital status was followed from 1942 to 2003, and cause-specific death rates and trends with exposure were evaluated. All cancers combined (standardized mortality ratio (SMR) = 1.0, 95% confidence interval (CI): 0.8, 1.1), lung cancers (SMR = 0.7, 95% CI: 0.5, 0.9), and nonmalignant respiratory disease (SMR = 0.8, 95% CI: 0.6, 1.0) were at or below expected levels. Observed deaths for leukemia (SMR = 1.9, 95% CI: 1.0, 3.2), non-Hodgkin lymphoma (SMR = 1.3, 95% CI: 0.6, 2.5), diabetes (SMR = 1.1, 95% CI: 0.6, 1.8), and ischemic heart disease (SMR = 1.1, 95% CI: 0.9, 1.2) were slightly greater than expected. No trend was observed with exposure for these causes of death. However, for 4 deaths of soft tissue sarcoma (SMR = 4.1, 95% CI: 1.1, 10.5), the mortality rates increased with exposure. The small number of deaths and the uncertainty in both diagnosis and nosology coding make interpretation of this finding tenuous. With the exception of soft tissue sarcoma, the authors found little evidence of increased disease risk from exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin.

Cigarette Smoking and Pancreatic Cancer: A Pooled Analysis From the Pancreatic Cancer Cohort Consortium

2009-06-26

Smoking is an established risk factor for pancreatic cancer; however, detailed examination of the association of smoking intensity, smoking duration, and cumulative smoking dose with pancreatic cancer is limited. The authors analyzed pooled data from the international Pancreatic Cancer Cohort Consortium nested case-control study (1,481 cases, 1,539 controls). Odds ratios and 95% confidence intervals were calculated by using unconditional logistic regression. Smoking intensity effects were examined with an excess odds ratio model that was linear in pack-years and exponential in cigarettes smoked per day and its square. When compared with never smokers, current smokers had a significantly elevated risk (odds ratio (OR) = 1.77, 95% confidence interval (CI): 1.38, 2.26). Risk increased significantly with greater intensity (≥30 cigarettes/day: OR = 1.75, 95% CI: 1.27, 2.42), duration (≥50 years: OR = 2.13, 95% CI: 1.25, 3.62), and cumulative smoking dose (≥40 pack-years: OR = 1.78, 95% CI: 1.35, 2.34). Risk more than 15 years after smoking cessation was similar to that for never smokers. Estimates of excess odds ratio per pack-year declined with increasing intensity, suggesting greater risk for total exposure delivered at lower intensity for longer duration than for higher intensity for shorter duration. This finding and the decline in risk after smoking cessation suggest that smoking has a late-stage effect on pancreatic carcinogenesis.

Age of Onset in Concordant Twins and Other Relative Pairs With Multiple Sclerosis

Sadovnick, A. D., Yee, I. M., Guimond, C., Reis, J., Dyment, D. A., Ebers, G. C. — 2009-06-22

The ages of onset in multiple sclerosis cases span more than 7 decades. Data are presented for affected relative pairs from a Canadian population base of 30,000 multiple sclerosis index cases (1993–2008). The effects of genetic sharing, parent of origin, intergenerational versus collinear differences, and gender on the ages of onset were evaluated in the following concordant pairs: monozygotic twins (n = 29), dizygotic twins (n = 10), siblings (n = 614), first cousins (n = 405), half siblings (n = 29), parent/child (n = 285), and aunt/uncle/niece/nephew (avunculars) (n = 289). Fisher's z test assessed intraclass correlation (r) for ages of onset. Correlations for monozygotic twins, dizygotic twins, full siblings, and first cousins were 0.60, 0.54, 0.20, and 0.10, respectively. Dizygotic twins resembled monozygotic twins more than siblings. The age-of-onset correlation for maternal half siblings (r = 0.37) was higher than that for paternal half siblings (r = 0.26), consistent with other observations suggesting an intrauterine environmental effect on multiple sclerosis risk. Intergenerational comparisons are complicated by substantial increases of multiple sclerosis incidence over time. Genetic loading (familial vs. sporadic cases) did not generally influence the age of onset, but correlation of age of onset in multiple sclerosis relative pairs was proportional to genetic sharing. A maternal parent-of-origin effect on the age of onset in collinear generations was suggested.

Hepatitis C Infection Among Injecting Drug Users in England and Wales (1992-2006): There and Back Again?

2009-06-22

Changes in hepatitis C virus (HCV) prevalence from 1992 to 2006 were examined by using 24,311 records from unlinked anonymous surveillance of injecting drug users in England and Wales. Bayesian logistic regression was used to estimate annual prevalence, accounting for changing recruitment patterns (age, gender, injecting duration, geographic region, interactions) and the sensitivity and specificity of different oral fluid testing devices. After controlling for these differences, the authors found that the adjusted HCV prevalence decreased from 70% (95% credible interval: 62, 78) in 1992 to 47% (95% credible interval: 43, 51) in 1998 before rising again to 53% (95% credible interval: 48, 58) in 2006. Women injecting drug users had a higher HCV risk than did men (odds ratio = 1.50, 95% credible interval: 1.31, 1.73). Two regions (London and North West) had a markedly higher HCV prevalence than did the rest of England and Wales. Among individuals who had injected for less than 1 year, the adjusted HCV prevalence in 2006 was higher than that in 1992 (28% vs. 19%, respectively). HCV infection can be prevented. The public health challenge in England and Wales is to increase action in order to regain a downward trend in HCV risk and the benefit that has been lost since 1998.

Cognitive Function and Overweight in Preschool Children

2009-06-22

The authors assessed the association between cognitive function and incidence and maintenance of overweight in preschool children. A population-based birth cohort was established in Menorca, Spain, between 1997 and 1999 (n = 482). Body mass index (weight (kg)/height (m)²) was measured at ages 4 years and 6 years (n = 421). At age 4 years, children were assessed for cognitive function (McCarthy Scales of Children's Abilities) (n = 395). After multivariable adjustment for a wide range of factors, including maternal education and body mass index, children with higher general cognition at age 4 years had a lower likelihood of being overweight (odds ratio = 0.47, 95% confidence interval (CI): 0.25, 0.88) at age 6 years. Children with higher general cognition at age 4 years had a lower likelihood of maintaining an unhealthy weight status (being at risk of overweight or overweight) between ages 4 years and 6 years, as well as worsening weight status over time, than children who maintained a healthy weight (odds ratios were 0.78 (95% CI: 0.54, 1.14) and 0.77 (95% CI: 0.51, 1.14), respectively). When specific dimensions of cognitive function were assessed, associations were mainly found for verbal and executive function areas. Children with higher cognitive function in early life might be at decreased risk of overweight later in childhood.

Ideal Weight and Weight Satisfaction: Association With Health Practices

Kuk, J. L., Ardern, C. I., Church, T. S., Hebert, J. R., Sui, X., Blair, S. N. — 2009-06-22

Evidence suggests that individuals have become more tolerant of higher body weights over time. To investigate this issue further, the authors examined cross-sectional associations among ideal weight, examination year, and obesity as well as the association of ideal weight and body weight satisfaction with health practices among 15,221 men and 4,126 women in the United States. Participants in 1987 reported higher ideal weights than participants in 2001, an effect particularly pronounced from 1987 to 2001 for younger and obese men (85.5 kg to 94.9 kg) and women (62.2 kg to 70.5 kg). For a given body mass index, higher ideal body weights were associated with greater weight satisfaction but lower intentions to lose weight. Body weight satisfaction was subsequently associated with greater walking/jogging, better diet, and lower lifetime weight loss but with less intention to change physical activity and diet or lose weight (P < 0.01). Conversely, body mass index was negatively associated with weight satisfaction (P < 0.01) and was associated with less walking/jogging, poorer diet, and greater lifetime weight loss but with greater intention to change physical activity and diet or lose weight. Although the health implications of these findings are somewhat unclear, increased weight satisfaction, in conjunction with increases in societal overweight/obesity, may result in decreased motivation to lose weight and/or adopt healthier lifestyle behaviors.

Contact With Beach Sand Among Beachgoers and Risk of Illness

2009-06-18

Recent studies of beach sand fecal contamination have triggered interest among scientists and in the media. Although evidence shows that beach sand can harbor high concentrations of fecal indicator organisms, as well as fecal pathogens, illness risk associated with beach sand contact is not well understood. Beach visitors at 7 US beaches were enrolled in the National Epidemiological and Environmental Assessment of Recreational Water (NEEAR) Study during 2003–2005 and 2007 and asked about sand contact on the day of their visit to the beach (digging in the sand, body buried in the sand). Then, 10–12 days after their visit, participants were telephoned to answer questions about any health symptoms experienced since the visit. The authors completed 27,365 interviews. Digging in the sand was positively associated with gastrointestinal illness (adjusted incidence proportion ratio (aIPR) = 1.13, 95% confidence interval (CI): 1.02, 1.25) and diarrhea (aIPR = 1.20, 95% CI: 1.05, 1.36). The association was stronger between those buried in the sand and gastrointestinal illness (aIPR = 1.23, 95% CI: 1.05, 1.43) and diarrhea (aIPR = 1.24, 95% CI: 1.01, 1.52). Nonenteric illnesses did not show a consistent association with sand contact activities. Sand contact activities were associated with enteric illness at beach sites. Variation in beach-specific results suggests that site-specific factors may be important in the risk of illness following sand exposure.

Life-Years Gained Among US Adults From Modern Treatments and Changes in the Prevalence of 6 Coronary Heart Disease Risk Factors Between 1980 and 2000

2009-06-18

Has the recent US decline in coronary heart disease (CHD) mortality increased life expectancy? The authors estimated the number of life-years gained from CHD treatments and changes in the prevalence of cardiovascular disease risk factors for the US population between 1980 and 2000. The previously validated IMPACT model was used to integrate data on numbers of CHD patients, treatment uptake, treatment effectiveness, population risk factor trends, and median survival among US adults. There were 308,900 fewer CHD deaths in 2000 among Americans aged 25–84 years than if 1980 mortality rates had applied. These 308,900 fewer deaths represented approximately 3,147,800 life-years gained (sensitivity analysis range, 2,448,900–3,744,900). Treatments for patients accounted for approximately 1,092,400 (751,700–1,387,000) life-years gained, whereas changes in the prevalence of population risk factors accounted for a gain of 2,055,500 (1,697,200–2,346,300) life-years. However, the 2,770,500 life-years gained through decreased levels of smoking, cholesterol, blood pressure, and physical inactivity were diminished by a loss of 715,000 life-years attributable to increased rates of obesity and diabetes. Therefore, modest reductions in the prevalence of several major cardiovascular disease risk factors accounted for more than twice as many life-years gained as did treatments. Unfortunately, these gains were partially offset by substantial increases in obesity and diabetes.

Alcohol and Risk of Breast Cancer by Histologic Type and Hormone Receptor Status in Postmenopausal Women: The NIH-AARP Diet and Health Study

2009-06-18

Little is known about the association between alcohol and breast cancer by different tumor characteristics. The study consisted of 184,418 postmenopausal women aged 50–71 years in the National Institutes of Health-AARP Diet and Health Study (1995–2003). Alcohol use, diet, and potential risk factors for cancer were assessed with a mailed questionnaire at baseline. The relative risks and 95% confidence intervals were estimated by using Cox proportional hazards regression. Breast cancer cases and estrogen receptor and progesterone receptor status were identified through linkage to state cancer registries. During an average of 7 years of follow-up, 5,461 breast cancer cases were identified. Alcohol was significantly positively associated with total breast cancer: Even a moderate amount of alcohol (>10 g/day) significantly increased breast cancer risk. In a comparison of >35 g versus 0 g/day, the multivariate relative risks were 1.35 (95% confidence interval (CI): 1.17, 1.56) for total breast cancer, 1.46 (95% CI: 1.22, 1.75) for ductal tumors, and 1.52 (95% CI: 0.95, 2.44) for lobular tumors. The multivariate relative risks for estrogen receptor-positive/progesterone receptor-positive, estrogen receptor-positive/progesterone receptor-negative, and estrogen receptor-negative/progesterone receptor-negative tumors were 1.46 (95% CI: 1.12, 1.91) for >35 g versus 0 g/day, 1.13 (95% CI: 0.73, 1.77) for >20 g versus 0 g/day, and 1.21 (95% CI: 0.79, 1.84) for >20 g versus 0 g/day, respectively. Moderate consumption of alcohol was associated with breast cancer, specifically hormone receptor-positive tumors.

Defining Incident Chronic Kidney Disease in the Research Setting: The ARIC Study

2009-06-17

Deaths of participants and losses to follow-up pose challenges for defining outcomes in epidemiologic studies. The authors compared several definitions of incident chronic kidney disease (CKD) in terms of incidence, agreement, and risk factor associations. They used data from 14,873 participants in the community-based, multicenter, biracial Atherosclerosis Risk in Communities Study (1987–1999). The estimated glomerular filtration rate (eGFR) was based on serum creatinine at baseline and the 3- and 9-year follow-up visits. Hospitalizations were ascertained continuously. The authors compared 4 definitions of incident CKD: 1) low eGFR (<60 mL/minute/1.73 m²); 2) low and declining (≥25%) eGFR; 3) an increase in serum creatinine (≥0.4 mg/dL) at 3- or 9-year follow-ups; and 4) CKD-related hospitalization or death. From these definitions, they identified 1,086, 677, 457, and 163 cases, respectively. There was relatively good agreement among definitions 1–3, but definition 4 identified mostly different cases. Risk factor associations were consistent across definitions for hypertension and lipids. Diabetes showed weaker associations with definition 1 (incidence rate ratio = 1.5, 95% confidence interval: 1.2, 1.7) than with definition 4 (incidence rate ratio = 6.3, confidence interval: 4.4, 8.9). Associations with gender differed in direction and magnitude across definitions. Case definition can impact relative risk estimates for CKD risk factors.

Invited Commentary: Defining Incident Chronic Kidney Disease in Epidemiologic Study Settings

Tonna, S. J. — 2009-06-17

Chronic kidney disease affects an estimated 31 million Americans and potentially poses a significant global health and socioeconomic crisis. Chronic kidney disease can be treated if patients are identified early enough in the evolution of their kidney disease. However, in order for this to occur, suitable definitions of what is meant by "chronic kidney disease" need to be identified. In clinical practice, prevalent chronic kidney disease is diagnosed in a patient on the basis of the presence of persistent albuminuria and/or reduced glomerular filtration rate. However, it is unclear how to best define an incident of chronic kidney disease when the definition relies on the need for a patient to be seen multiple times over an extended period of time. In this issue of the Journal, Bash et al. (Am J Epidemiol. 2009;170(0):000–000) have compared 4 different definitions of incident chronic kidney disease and their agreement, incident rates, and association with known risk factors. This study explores an extremely important topic for longitudinal epidemiology studies of chronic kidney disease.

Growth Trajectories and Intellectual Abilities in Young Adulthood: The Helsinki Birth Cohort Study

2009-06-15

Slow childhood growth is associated with poorer intellectual ability. The critical periods of growth remain uncertain. Among 2,786 Finnish male military conscripts (1952–1972) born in 1934–1944, the authors tested how specific growth periods from birth to age 20 years predicted verbal, visuospatial, and arithmetic abilities at age 20. Small head circumference at birth predicted poorer verbal, visuospatial, and arithmetic abilities. The latter 2 measures were also associated with lower weight and body mass index (weight (kg)/height (m)²) at birth (for a 1-standard-deviation (SD) decrease in test score per SD decrease in body size ≥ 0.05, P’s < 0.04). Slow linear growth and weight gain between birth and age 6 months, between ages 6 months and 2 years, or both predicted poorer performance on all 3 tests (for a 1-SD decrease in test score per SD decrease in growth ≥ 0.05, P’s < 0.03). Reduced linear growth between ages 2 and 7 years predicted worse verbal ability, and between age 11 years and conscription it predicted worse performance on all 3 tests. Prenatal brain growth and linear growth up to 2 years after birth form a first critical period for intellectual development. There is a second critical period, specific for verbal development, between ages 2 and 7 years and a third critical period for all 3 tested outcomes during adolescence.

THE PRIMARY AUTHORS REPLY

Lin, S., Munsie, J. P. W. — 2009-06-10

RE: "MATERNAL ASTHMA MEDICATION USE AND THE RISK OF GASTROSCHISIS"

Payne, N. R., Mijal, R. S. — 2009-06-10

How Accurately Do Young Adults Recall Childhood Pets? A Validation Study

Nicholas, C., Wegienka, G., Havstad, S., Ownby, D., Johnson, C. C., Zoratti, E. — 2009-06-10

Epidemiologic research shows that pets influence human health, demonstrating both protective and deleterious health risks; therefore, valid definitions of pet exposure would enhance research. The authors determined how well young adults aged 18 years report their early childhood pets. Subjects in an established birth cohort from Detroit, Michigan, born in 1987–1989 (n = 820) were asked a series of questions about pets in the home during their first 6 years of life. Pet recall was compared with annual prospectively collected parental report from 12–18 years prior. Exposure to cats was correctly reported on average 86.3% of the time (95% confidence interval: 85.0, 87.5) and dogs 79.2% (95% confidence interval: 77.7, 80.6) of the time (P < 0.01). Cats and dogs were more likely to be underreported than overreported, from as few as 1.8-fold to as many as 8.3-fold (P < 0.05). Reporting differed by sex of the respondent and current pet ownership. No differences were found in reporting by those who experienced allergy symptoms near dogs or cats. Findings suggest good reliability of young adult pet reporting for ages 0–6 years but that childhood pet exposure may need to be assessed separately depending on the participant's sex and the outcome of interest.

A Systematic Evaluation of Factors Associated With Nocturia--The Population-based FINNO Study

2009-06-10

In a case-control study with prevalence sampling, the authors explored the correlates for nocturia and their population-level impact. In 2003–2004, questionnaires were mailed to 6,000 subjects (aged 18–79 years) randomly identified from the Finnish Population Register (62.4% participated; 53.7% were female). Questionnaires contained items on medical conditions, medications, lifestyle, sociodemographic and reproductive factors, urinary symptoms, and snoring. Nocturia was defined as ≥2 voids/night. In age-adjusted analyses, factors associated with nocturia were entered into a multivariate model. Backward elimination was used to select variables for the final model, with adjustment for confounding. Although numerous correlates were identified, none affected ≥50% of nocturia cases of both sexes. The factors with the greatest impact at the population level were (urinary) urgency (attributable number/1,000 subjects (AN) = 24), benign prostatic hyperplasia (AN = 19), and snoring (AN = 16) for men and overweight and obesity (AN = 40), urgency (AN = 24), and snoring (AN = 17) for women. Moreover, correlates included prostate cancer and antidepressant use for men, coronary artery disease and diabetes for women, and restless legs syndrome and obesity for both sexes. Although several correlates were identified, none accounted for a substantial proportion of the population burden, highlighting the multifactorial etiology of nocturia.

Placental Vascular Pathology Findings and Pathways to Preterm Delivery

2009-06-09

The authors examined the associations between placental vascular findings and preterm delivery in 1,053 subcohort women (239 preterm, 814 term) from a Michigan pregnancy cohort study (1998–2004). Twenty-nine placental vascular variables from microscopic examinations were grouped into 5 constructs: 3 maternal constructs—obstructive lesions (MV-O), bleeding/vessel integrity (MV-I), and lack of physiologic conversion of maternal spiral arteries (MV-D)—and 2 fetal constructs—obstructive lesions (FV-O) and bleeding/vessel integrity (FV-I). Construct-specific scores were created by adding the number of positive findings and deriving a dichotomous variable to approximate the top quintile ("high") and bottom 4 quintiles ("not high") within each construct. In multivariate polytomous logistic regression models, medically indicated preterm delivery at <35 weeks was significantly associated with high scores for each of the vascular constructs; adjusted odds ratios ranged from 2.4 to 5.4. Spontaneous preterm delivery at 35–36 weeks was significantly associated with a high score on any 1 of 3 constructs: MV-I, MV-D, and FV-I. Spontaneous preterm delivery at <35 weeks was significantly associated with a high score on 2 or more of 3 constructs: MV-I, MV-D, and FV-I; adjusted odds ratios ranged from 4.1 to 7.4. These results support a role for various placental vascular lesions in medically indicated and spontaneous preterm delivery.

Holzman et al. Respond to "Intrauterine Epidemiology"

Holzman, C., Senagore, P., Kelly, R. — 2009-06-09

Invited Commentary: Intrauterine Epidemiology

McElrath, T. F., Hecht, J. L. — 2009-06-09

Traditionally, the investigation of preterm birth has relied on diagnostic definitions derived from maternal clinical presentation. However, clinical presentation may be only tangentially related to the underlying etiology of a disease. The utilization of data derived directly from the intrauterine or maternal systemic environment would be invaluable in consideration of the causes of preterm birth. In this issue, Kelly et al. (Am J Epidemiol. 2009;000(00)000–000) contribute to our understanding of the epidemiology of the intrauterine environment by classifying the vascular biology of the maternal-placental interface in cases of preterm delivery. Their histology-based approach observes that vascular conditions may be grouped into 5 constructs with specific relations to maternal and fetal vascular pathology. The frequencies of these constructs vary with regard to delivery indication and gestational age, suggesting that the intrauterine conditions associated with preterm birth are more complicated than originally appreciated. This work is laborious, and replication of the technique will be important. However, these authors have taken a large step toward introducing an "intrauterine" perspective into perinatal epidemiology and into our understanding of the underlying etiologies of preterm birth.

Designing 2-Phase Prevalence Studies in the Absence of a "Gold Standard" Test

Salim, A., Welsh, A. H. — 2009-06-08

A population survey for estimating prevalence is challenging when a disease or condition is difficult to diagnose. If clinical diagnosis is expensive, a 2-phase study, in which less expensive but less accurate tests are administered to all study subjects in the first phase (screening phase) and a more accurate but expensive or time-consuming test is administered to only a subset of the subjects in the second phase, is an attractive approach. Published research has discussed ways of maximizing precision of the prevalence estimate from a 2-phase study with a "gold standard" second-phase test. For many psychiatric disorders, even the best diagnostic tests are not of gold standard quality. In this paper, the authors propose a quasi-optimal design for 2-phase prevalence studies without a gold standard test; random-effects latent class analysis facilitates the estimation of prevalence and appropriately addresses the issue of dependent errors among the diagnostic tests. The authors show that the quasi-optimal design is efficient compared with the balanced and random designs when there is strong inter-test dependence caused by additional factors, apart from disease status, and highlight the importance of collecting data on those subjects testing negative in the first phase.

Duration of Antidepressant Drug Treatment and Its Influence on Risk of Relapse/Recurrence: Immortal and Neglected Time Bias

Gardarsdottir, H., Egberts, T. C., Stolker, J. J., Heerdink, E. R. — 2009-06-04

Several observational studies have found a higher risk of recurrence/relapse of depression for patients who discontinue antidepressant use compared with those who continue. This study demonstrated that measurement of follow-up time can be subject to immortal and neglected time bias. Data were obtained from the 2001 Second Dutch National Survey of General Practice. The study population was composed of antidepressant users with a registered depression diagnosis, divided into early discontinuers and continuing users. Two methods were used to measure time to relapse/recurrence. Method 1, used in previously mentioned studies, measured the beginning of follow-up 6 months after starting antidepressant therapy. Method 2 constructed individual treatment episodes for each patient and measured follow-up from actual end-of-treatment episode. The Cox proportional hazards model produced a risk ratio of 1.58 (95% confidence interval: 1.02, 2.45) for method 1, suggesting a higher risk of relapse/recurrence for early discontinuers. In method 2, a statistically nonsignificant risk ratio of 0.77 (95% confidence interval: 0.49, 1.21) was produced, indicating no difference in risk of relapse/recurrence. The authors found the method used in previous studies subject to bias. Applying a different method, accounting for immortal and neglected time bias, eliminated the protective effects of longer treatments.

Gardarsdottir et al. Respond to "Assessing Treatment Effects With Observational Analyses"

Gardarsdottir, H., Egberts, T. C., Stolker, J. J., Heerdink, E. R. — 2009-06-04

Invited Commentary: Assessing Treatment Effects by Using Observational Analyses--Opportunities and Limitations

Wang, P. S., Schoenbaum, M. — 2009-06-04

Making decisions about medical treatments based upon valid evidence is critical to improve health-care quality, outcomes, and value. Although such research commonly connotes the use of randomized controlled trials, experimental methods are not always feasible, and research using observational, quasi-experimental, and other nonexperimental methods may also be important. At the same time, nonexperimental methods are inherently susceptible to various types of bias and thus present special challenges in the search for valid and generalizable evidence. The study by Gardarsdottir et al. (Am J Epidemiol. 2009;000(00):000–000), on which this commentary is based, addresses a key potential source of bias—mismeasurement of patients’ duration of treatment—in previous research on pharmacotherapy for depression. However, the authors’ study is unlikely to address other potential sources of bias, which may make interpretation of their findings more difficult.

Comparability and Representativeness of Control Groups in a Case-Control Study of Infant Leukemia: A Report From the Children's Oncology Group

2009-06-04

Traditionally, controls in US pediatric cancer studies were selected through random digit dialing. With declining participation and lack of nonparticipant information, random digit dialing (RDD) controls may be substandard. Birth certificate (BC) controls are an alternative, because they are population based and include data from nonparticipants. The authors examined controls collected by random digit dialing and birth certificates for a Children's Oncology Group case-control study of infant leukemia in 1995–2006. Demographic variables were used to assess differences in RDD and BC controls and their representativeness. RDD and BC controls did not differ significantly with regard to maternal variables (age, race, education, marital status, alcohol during pregnancy) or child variables (sex, gestational age, birth weight), but they varied in smoking during pregnancy (22% RDD controls, 12% BC controls) (P = 0.05). The study's combined control group differed significantly from US births: Mothers of controls were more likely to be older (29.8 vs. 27.2 years), white (84% vs. 59%), and married (85% vs. 67%) and to have >16 years of education (37% vs. 25%). Control children were more often full term (88% vs. 81%) and heavier (3,436 vs. 3,317 g). Finally, participating BC mothers were likely to be older and to have more education than nonparticipants. Thus, the study's control groups were comparable but differed from the population of interest.

Genome-Wide Association Studies, Field Synopses, and the Development of the Knowledge Base on Genetic Variation and Human Diseases

2009-06-04

Genome-wide association studies (GWAS) have led to a rapid increase in available data on common genetic variants and phenotypes and numerous discoveries of new loci associated with susceptibility to common complex diseases. Integrating the evidence from GWAS and candidate gene studies depends on concerted efforts in data production, online publication, database development, and continuously updated data synthesis. Here the authors summarize current experience and challenges on these fronts, which were discussed at a 2008 multidisciplinary workshop sponsored by the Human Genome Epidemiology Network. Comprehensive field synopses that integrate many reported gene-disease associations have been systematically developed for several fields, including Alzheimer's disease, schizophrenia, bladder cancer, coronary heart disease, preterm birth, and DNA repair genes in various cancers. The authors summarize insights from these field synopses and discuss remaining unresolved issues—especially in the light of evidence from GWAS, for which they summarize empirical P-value and effect-size data on 223 discovered associations for binary outcomes (142 with P < 10^–7). They also present a vision of collaboration that builds reliable cumulative evidence for genetic associations with common complex diseases and a transparent, distributed, authoritative knowledge base on genetic variation and human health. As a next step in the evolution of Human Genome Epidemiology reviews, the authors invite investigators to submit field synopses for possible publication in the American Journal of Epidemiology.

Competing Risk Regression Models for Epidemiologic Data

Lau, B., Cole, S. R., Gange, S. J. — 2009-06-03

Competing events can preclude the event of interest from occurring in epidemiologic data and can be analyzed by using extensions of survival analysis methods. In this paper, the authors outline 3 regression approaches for estimating 2 key quantities in competing risks analysis: the cause-specific relative hazard (_csRH) and the subdistribution relative hazard (_sdRH). They compare and contrast the structure of the risk sets and the interpretation of parameters obtained with these methods. They also demonstrate the use of these methods with data from the Women's Interagency HIV Study established in 1993, treating time to initiation of highly active antiretroviral therapy or to clinical disease progression as competing events. In our example, women with an injection drug use history were less likely than those without a history of injection drug use to initiate therapy prior to progression to acquired immunodeficiency syndrome or death by both measures of association (_csRH = 0.67, 95% confidence interval: 0.57, 0.80 and _sdRH = 0.60, 95% confidence interval: 0.50, 0.71). Moreover, the relative hazards for disease progression prior to treatment were elevated (_csRH = 1.71, 95% confidence interval: 1.37, 2.13 and _sdRH = 2.01, 95% confidence interval: 1.62, 2.51). Methods for competing risks should be used by epidemiologists, with the choice of method guided by the scientific question.

Using Cases and Parents to Study Multiplicative Gene-by-Environment Interaction

Kistner, E. O., Shi, M., Weinberg, C. R. — 2009-05-29

With case-parent triads, one can estimate genotype relative risks by measuring the apparent overtransmission of susceptibility genotypes from parents to affected offspring. Results obtained using such designs, properly analyzed, resist both bias due to population structure and bias due to self-selection. Most diseases are not purely genetic, and environmental cofactors can also be important. In this paper, the authors describe how a polytomous logistic regression method previously developed for studying genetic effects on a quantitative trait can be used with case-parent data to study multiplicative gene-by-environment interaction. The idea is that if the joint effect of exposure and genotype on risk is submultiplicative or supermultiplicative, then, conditional on the parental genotypes, inheritance of a susceptibility genotype by affected offspring will appear to have been influenced by the offspring's exposure level. The authors' approach tolerates exposure-complicated genetic population structure, and simulations suggest power and Type I error rates comparable to those of competitors. With this approach, one can estimate the usual interaction parameters under a much less stringent assumption than gene-environment independence in the source population. Incompletely genotyped triads can contribute through an expectation-maximization algorithm. To illustrate, the authors consider polymorphisms in detoxification pathway genes and maternal smoking in relation to the birth defect oral cleft.

Determinants of Quality of Interview and Impact on Risk Estimates in a Case-Control Study of Bladder Cancer

2009-05-28

The authors evaluated potential determinants of the quality of the interview in a case-control study of bladder cancer and assessed the effect of the interview quality on the risk estimates. The analysis included 1,219 incident bladder cancer cases and 1,271 controls recruited in Spain in 1998–2001. Information on etiologic factors for bladder cancer was collected through personal interviews, which were scored as unsatisfactory, questionable, reliable, or high quality by the interviewers. Eight percent of the interviews were unsatisfactory or questionable. Increasing age, lower socioeconomic status, and poorer self-perceived health led to higher proportions of questionable or unreliable interviews. The odds ratio for cigarette smoking, the main risk factor for bladder cancer, was 6.18 (95% confidence interval: 4.56, 8.39) overall, 3.20 (95% confidence interval: 1.13, 9.04) among unsatisfactory or questionable interviews, 6.86 (95% confidence interval: 4.80, 9.82) among reliable interviews, and 7.70 (95% confidence interval: 3.64, 16.30) among high-quality interviews. Similar trends were observed for employment in high-risk occupations, drinking water containing elevated levels of trihalomethanes, and use of analgesics. Higher quality interviews led to stronger associations compared with risk estimation that did not take the quality of interview into account. The collection of quality of interview scores and the exclusion of unreliable interviews probably reduce misclassification of exposure in observational studies.

Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating: By Ewout W. Steyerberg

Alonzo, T. A. — 2009-05-28

Body Mass Index, Height, and Risk of Lymphatic Malignancies: A Prospective Cohort Study

2009-05-28

The association among body mass index (BMI), height, and the risk of lymphatic malignancies was investigated in the Netherlands Cohort Study. The participants (n = 120,852), Dutch men and women aged 55–69 years, completed a self-administered questionnaire at baseline in 1986. After 13.3 years of follow-up, data on 1,042 lymphatic malignancy cases (including diffuse large-cell lymphoma, chronic lymphocytic leukemia, multiple myeloma, and other subtypes) and 4,588 subcohort members were available. Incidence rate ratios were estimated by using Cox regression models. BMI at baseline and BMI change since the age of 20 years were not associated with lymphatic malignancy risk. However, the rate ratio of lymphatic malignancies per 4-unit increase in BMI at 20 years of age was 1.13 (95% confidence interval (CI): 1.01, 1.25). The rate ratio of lymphatic malignancies per 5-cm increase in height was 1.08 (95% CI: 1.02, 1.15). For diffuse large-cell lymphoma, follicular lymphoma, and chronic lymphocytic leukemia, the relative risks were 1.19 (95% CI: 1.07, 1.33), 1.15 (95% CI: 0.95, 1.40), and 1.09 (95% CI: 0.95, 1.26), respectively, for each 5-cm increase in height. The positive associations among BMI at 20 years of age, height, and the risk of lymphatic malignancies suggest that exposures during early life play a role in the etiology of lymphatic malignancies.

Fetal Growth and Risk of Childhood Acute Lymphoblastic Leukemia: Results From an Australian Case-Control Study

2009-05-28

The relation between intrauterine growth and risk of childhood acute lymphoblastic leukemia was investigated in an Australian population-based case-control study that included 347 cases and 762 controls aged <15 years recruited from 2003 to 2006. Information on proportion of optimal birth weight, a measure of the appropriateness of fetal growth, was collected from mothers by questionnaire. Data were analyzed by using logistic regression. Risk of acute lymphoblastic leukemia was positively associated with proportion of optimal birth weight; the odds ratio for a 1-standard-deviation increase in proportion of optimal birth weight was 1.18 (95% confidence interval: 1.04, 1.35) after adjustment for the matching variables and potential confounders. This association was also present among children who did not have a high birth weight, suggesting that accelerated growth, rather than high birth weight per se, is associated with risk of acute lymphoblastic leukemia. Similar associations between proportion of optimal birth weight and acute lymphoblastic leukemia were observed for both sexes and across age groups and leukemia subtypes. Results of this study confirm earlier findings of a positive association between rapidity of fetal growth and subsequent risk of acute lymphoblastic leukemia in childhood, and they are consistent with a role for insulin-like growth factors in the causal pathway.

Dynamics and Control of Infections Transmitted From Person to Person Through the Environment

Li, S., Eisenberg, J. N. S., Spicknall, I. H., Koopman, J. S. — 2009-05-27

The environment provides points for control of pathogens spread by food, water, hands, air, or fomites. These environmental transmission pathways require contact network formulations more realistically detailed than those based on social encounters or physical proximity. As a step toward improved assessment of environmental interventions, description of contact networks, and better use of environmental specimens to analyze transmission, an environmental infection transmission system model that describes the dynamics of human interaction with pathogens in the environment is presented. Its environmental parameters include the pathogen elimination rate, µ, and the rate humans pick up pathogens, , and deposit them, . The ratio, N/µ (N equals population size), indicates whether transmission is density dependent (low ratio), frequency dependent (high ratio), or in between. Transmission through frequently touched fomites, such as doorknobs, generates frequency-dependent patterns, while transmission through thoroughly mixed air or infrequently touched fomites generates density-dependent patterns. The environmental contamination ratio, /, reflects total agent deposition per infection and outbreak probability, where is defined as the recovery rate. These insights provide theoretical contexts to examine the role of the environment in pathogen transmission and a framework to interpret environmental data to inform environmental interventions.

Antioxidant Enzyme Activity and Coronary Heart Disease: Meta-analyses of Observational Studies

2009-05-22

Controversial data exist concerning the relation between the activities of scavenger antioxidant enzymes and coronary heart disease (CHD) risk. The authors report updated meta-analyses of studies assessing the activities of 3 antioxidant enzymes—glutathione peroxidase, superoxide dismutase, and catalase—and CHD risk. Computer-based and manual searches of the relevant literature from January 1966 to January 2008 were performed. Studies assessing glutathione peroxidase, superoxide dismutase, and catalase activities in cells or biologic fluids and clinical CHD outcomes were selected. Pooled odds ratios for CHD were calculated by using an inverse-variance-weighted random-effects model. Forty-two case-control studies and 3 prospective studies were included. The pooled odds ratios for CHD associated with a 1-standard-deviation increase in glutathione peroxidase, superoxide dismutase, and catalase activity levels were 0.51 (95% confidence interval: 0.35, 0.75), 0.48 (95% confidence interval: 0.32, 0.72), and 0.32 (95% confidence interval: 0.16, 0.61), respectively, with substantial between-study heterogeneity (I² > 90% for the 3 enzymes). These findings were remarkably robust in the sensitivity analysis. The authors’ meta-analyses support an inverse association between circulating levels of superoxide dismutase, glutathione peroxidase, and catalase activities with CHD and emphasize the need for additional high-quality prospective studies.

Sleep-disordered Breathing and Frailty in the Cardiovascular Health Study Cohort

Endeshaw, Y. W., Unruh, M. L., Kutner, M., Newman, A. B., Bliwise, D. L. — 2009-05-22

Sleep-disordered breathing (SDB) is associated with pathophysiology that may influence the development and progression of frailty. Using data collected in 1995–1996, the authors explored the relation between SDB and components of frailty among 1,042 participants of the Cardiovascular Health Study. Diagnosis of SDB was based on the results of overnight polysomnography, and severe SDB was defined as an apnea-hypopnea index of >30 per hour of sleep. Slow walking speed, low grip strength, exhaustion, low physical activity, and unexplained weight loss were referred to as frailty indicator variables. There were 584 (56%) female and 458 (44%) male participants, and the mean age was 77 (standard deviation, 4) years. There was independent association between severe SDB and 1 or more frailty indicator variables (adjusted odds ratio = 4.85, 95% confidence interval: 1.40, 16.78), slow walking speed (adjusted odds ratio = 2.67, 95% confidence interval: 1.04, 6.84), and low grip strength (adjusted odds ratio = 3.29, 95% confidence interval: 1.36, 7.96) among female study participants. The finding of an independent association between SDB and frailty indicator variables among older women could have important implications in interventions aimed at preventing or delaying the progression of frailty.

Psychosocial Stress and Change in Weight Among US Adults

Block, J. P., He, Y., Zaslavsky, A. M., Ding, L., Ayanian, J. Z. — 2009-05-22

The association of psychosocial stress with weight gain may have important implications for clinical practice and workplace and public health interventions. To determine whether multiple domains of psychosocial stress were associated with weight gain from 1995 to 2004, the authors analyzed a nationally representative longitudinal cohort of 1,355 men and women in the United States. Change in body mass index was assessed for multiple domains of psychosocial stress related to work, personal relationships, life constraints, and finances, controlling for other factors associated with weight gain. All analyses were stratified by sex and weighted to account for the complex survey design. Among men with high baseline body mass index, weight gain was associated with increasing levels of psychosocial stress related to job-related demands (P < 0.001 for interaction with baseline body mass index), lack of skill discretion (P = 0.014), lack of decision authority (P = 0.026), and difficulty paying bills (P = 0.004). Among women with high baseline body mass index, weight gain was associated with job-related demands (P < 0.001 for interaction with baseline body mass index), perceived constraints in life (P < 0.001), strain in relations with family (P = 0.016), and difficulty paying bills (P = 0.010). Interventions to address psychosocial stress may limit weight gain among overweight and obese men and women.

Caregiving Intensity and Change in Physical Functioning Over a 2-Year Period: Results of the Caregiver-Study of Osteoporotic Fractures

Fredman, L., Doros, G., Ensrud, K. E., Hochberg, M. C., Cauley, J. A. — 2009-05-14

It is unknown whether caregivers who perform more caregiving tasks have a greater decline in health from higher stress or less decline because of better health, staying active, or psychological factors. This 1999–2004 US study examined caregiving intensity and 2-year change in performance-based functioning among 901 elderly women from the Caregiver-Study of Osteoporotic Fractures sample. Caregivers were categorized as high (n = 167) or low (n = 166) intensity based on how many activities of daily living they performed for the care recipient. Caregiving intensity status and physical performance score (sum of quartiles of walking pace, grip strength, and chair-stand speed; range, 0–9) were assessed at baseline and at 2 annual follow-up interviews. At baseline, high-intensity caregivers reported the most stress but had the best physical functioning; noncaregivers (n = 568) had the poorest physical functioning (adjusted scores = 5.09 vs. 4.54, P = 0.03). Low-intensity caregivers declined more than noncaregivers over 2 years, but high-intensity caregivers did not (adjusted difference = –0.33, P = 0.07 vs. 0.03, P = 0.89). Among respondents with the same caregiving status at baseline and 1-year interviews, high-intensity caregivers maintained the highest physical performance throughout follow-up. Higher levels of physical performance persisted over 2 years among high-intensity caregivers, which did not support the traditional stress hypothesis.

Effect of Reducing Indoor Air Pollution on Women's Respiratory Symptoms and Lung Function: The RESPIRE Randomized Trial, Guatemala

2009-05-14

Exposure to household wood smoke from cooking is a risk factor for chronic obstructive lung disease among women in developing countries. The Randomized Exposure Study of Pollution Indoors and Respiratory Effects (RESPIRE) is a randomized intervention trial evaluating the respiratory health effects of reducing indoor air pollution from open cooking fires. A total of 504 rural Mayan women in highland Guatemala aged 15–50 years, all using traditional indoor open fires, were randomized to either receive a chimney woodstove (plancha) or continue using the open fire. Assessments of chronic respiratory symptoms and lung function and individual measurements of carbon monoxide exposure were performed at baseline and every 6 months up to 18 months. Use of a plancha significantly reduced carbon monoxide exposure by 61.6%. For all respiratory symptoms, reductions in risk were observed in the plancha group during follow-up; the reduction was statistically significant for wheeze (relative risk = 0.42, 95% confidence interval: 0.25, 0.70). The number of respiratory symptoms reported by the women at each follow-up point was also significantly reduced by the plancha (odds ratio = 0.7, 95% confidence interval: 0.50, 0.97). However, no significant effects on lung function were found after 12–18 months. Reducing indoor air pollution from household biomass burning may relieve symptoms consistent with chronic respiratory tract irritation.

Associations of Gestational Weight Gain With Short- and Longer-term Maternal and Child Health Outcomes

Oken, E., Kleinman, K. P., Belfort, M. B., Hammitt, J. K., Gillman, M. W. — 2009-05-13

The authors investigated the rate of gestational weight gain associated with the lowest combined risk of 5 short- and longer-term maternal and child health outcomes for 2,012 mother-child pairs recruited in 1999–2002 into Project Viva, a prebirth cohort study in Massachusetts. Within each maternal prepregnancy body mass index (BMI, kg/m²) stratum, they performed a logistic regression analysis predicting all 5 outcomes, from which they determined the rate of gain at which average predicted prevalence of the adverse outcomes was the lowest. The mean rate of total gestational weight gain was 0.39 kg/week (standard deviation, 0.14). The prevalence of small for gestational age was 6%, large for gestational age was 14%, preterm delivery was 7%, substantial postpartum weight retention was 16%, and child obesity was 10%. The lowest predicted outcome prevalence occurred with a 0.28-kg/week gain for women whose BMI was 18.5–24.9, a 0.03-kg/week loss for a BMI of 25.0–29.9, and a 0.19-kg/week loss for a BMI of ≥30.0 kg/m²—the lowest observed weight changes in overweight and obese women. For normal-weight and overweight women, lowest-risk gains varied modestly with adjustment for maternal characteristics and with different outcome weightings. For obese women, the lowest-risk weight change was weight loss in all models. Recommendations for gestational weight gain for obese women should be revised.